Monica Turner | Ecosystems and the Ecology of Change

So I’m gonna be discussing changing
ecosystems and the ecology of change but at the outset, I promise there was
no sharing of slides ahead of time there’ll be a little redundancy in my first few
slides here, but I would like to say congratulations to the whole Odum School.
I mean it’s just incredible to think that it’s been 50 years already in the
making and here we are today. The Institute as has been mentioned
previously really was a vision ahead of its time and the reason that we are
where we are is that there has been effective leadership all along the way. As we heard earlier institutional support from UGA, the Institute wouldn’t
be where it is without that and we had tremendously committed faculty, many of
whom are here today who devoted their entire careers and and grew as
scientists at this place. So it’s a really remarkable accomplishment. So I
did just do a little bit of walking down memory lane. All of my pictures are
slides or photos which I have photographed but the Institute when I
first started so this is my first year, February 1981. You will notice an absence
of vegetation around that building it’s strikingly different. Those of you who
were — how many of you were in the Carroll’s? Your offices in the Carroll’s?
A number of you. Right, so this is these are the Carroll’s this is my side of the
building not the other side where Kim and Bob were. That’s my desk. I wish my
desk was as cleaned today as it is then. Our courtyard, I mean that was such
a wonderful place to read our papers for class. I mean taking your
paper out and sitting out in the courtyard or having potlucks. And then
again the absence of –whoopsy [computer alert] Here we go. Gotta get the pointer right. That’s where
we used to play volleyball on Fridays and drink beer so things have changed
again as well but what I really value so much in my own training and the
experience I had here was that we had such leading thinkers present training
us as young scientists, as early career scientists, and representing the science
of the times. That’s my copy of Odum’s Ecology book which I had as an
undergraduate. It’s why I applied to the University of Georgia. As in New Yorker
going to Georgia was not considered to be the best thing from my family members. It was awesome however but we were lucky to have so
many people leading in so many areas. I have names here and if there are any I
omitted it’s simply by a lack of remembering it’s not my intention, but I
mean Gene Odum and Frank for ecosystem ecology. Sustainability a term now that
was that was new then. Systems ecology and modeling: Bernie Patten and Dick
Weigert, long-term research: so D.A.C. , Wayne Swank at Coweta, and Judy. Biogeochemical cycling: Bruce Haines, Carl Jordan, Microbial loop which was new in marine systems and marine ecology: Larry Pomeroy, Jim Porter, Aquatic ecology: Judy Meyer, Karen Porter especially Agroecology: was
new D.A.C. and Dave Coleman were leading a lot of that effort and then landscape ecology: Frank Golley, Vernon Meentemeyer in geography and Ron Pulliam with a lot
of his source sink ideas. As was mentioned previously, these areas and
these leaders who are pioneering in their thinking. They were holistic. They
were integrative and collaborative. Now not only were they leading thinkers but
they were thoughtful leaders, leading us by example. This was mentioned previously
but Georgia has five prior ESA Presidents among its faculty. That is the
only institution to have that. We grew up, as young scientists, being trained to
contribute back to our discipline to help serve the societies upon which we
depend and from which we benefit. Frank was also leading a lot of the
International connections by his service as as president of INTECOL. And also I
feel like it’s appropriate for me to give a couple of personal notes of
gratitude to some of these faculty who influenced me so much of my development.
Frank was my mentor, my adviser. He was extraordinary as a mentor for a young scientist and I would not be where I am today. I would not have evolved in my
science without that that training. Susan Bratton who’s pictured here was
the person who taught me field ecology. She was on my committee. She was from the
Park Service co-op unit. They funded my research as well. But as a female
scientist, she along with Judy and Karen There WERE women scientists
here that was unusual. There were few in the ranks of the faculty so I am greatly
appreciative for both Susan on my committee but the others as well
Gene was on my committee he also hired me as a postdoc when I had no other
offers but it was also a wonderful opportunity. Larry Pomeroy accepted me as a student, which many of you may not know, I had
done my undergraduate honors thesis on phytoplankton limitations — limitation of
silica on phytoplankton in Long Island Sound and unfortunately I learned I
didn’t want to sit by a microscope so I got here naively and said I don’t really
want to do that and he was very gracious D.A.C. was head of Graduate Admissions and
was really really just delightful in helping to recruit me. And Bernie
Patton I would call out as well. The course on systems ecology, I don’t know
how many of you have had Bernie’s class in systems, quite a number of you. It was highly influential and really changed my thinking and in a good way So I’m going
to be talking today though about science. And I will be talking about the ecology
of change. I will mention however that UGA has continually evolved with the
times after 50 years it’s still at the forefront addressing many of the
contemporary problems we face in ecology but as has been mentioned by Peter and
others, we’re currently in a time of just unprecedented rates of change and some
of the changes that we are observing and in all of the systems in which we
work are big. That is there qualitative differences and they happen really fast. They’re abrupt changes and I think these are all things that lead to surprising
outcomes where we say “Geez, we didn’t expect that!” And I think it’s also one of
our biggest challenges in contemporary ecology. Regardless of the level of
biological organization or a sub discipline or the scale . So that’s going
to be kind of my theme again following up a little bit on the rates of change
that Peter so nicely set up. This is from Will Steffen’s paper updating some of the
big drivers globally have change. There’s socio-economics on this side, Earth
system trends on the other. Details don’t – aren’t critical but what I’d like to
just focus your attention is since 1950, that exponential, that hockey stick
curve that you see in so many different drivers. So this is fundamentally
changing the way our ecological systems are responding and I apologize for the
the acronym here ahead of identifying it but A.C.E.S. standing for abrupt change in
ecological systems are happening all around us the more we look for them, the
more we see them. And they’re continuing to surprise. So whether it’s lakes
suddenly turning from clear to turbid or eutrophic or coral reefs that are going
from these beautiful colorful diverse systems into areas where they’re dying
or rangelands for example where we’re seeing increased densification of woody
vegetation and bareness on the ground from what had been a a thriving
rangeland system. So abrupt change when I talk about that. It’s things that are
abrupt in time or relative to the speed at which the drivers are changing. If you
think about some response, a driver or a state you think about time. It could be a
step function but it’s big and it’s happening quickly. It could be a little
bit, slightly different sorts of changes but we’re talking about big changes that
happen quickly. Oftentimes these involve tipping points or thresholds. Where again,
if we have our response or our state on the Y and we have some kind of
controlling driver on the X where if you’re near that threshold, a very slight
change in the value of that driver, can cause a really large change in
response. It may be a threshold. It may also be a hysteretic or bifurcation
alternative state type dynamic. Where when you release the driver, you push it
back the other way, you can’t just get the system to recover quite well. It’s
very difficult in practice to distinguish those two. And I would say
one of our big picture questions in ecology is ‘ when where and why are these
kinds of changes going to be happening in our systems? Can we anticipate them?
Can we figure out ways to avoid the ones that we don’t want? Can we diagnose them
when they happen? So, we’re actually involved currently in a project that’s
focusing on this general area across systems, timescales, spatial scales,
terrestrial, aquatic. A group of five faculty and a cluster hire of four
postdocs so we’ve been trying to think through this very generally and it’s
really challenging and surprisingly so actually once we once we get into it. So, the abrupt changes are hard to diagnose because of several factors. One
is they can be caused by many different causes – that didn’t make sense –
They can arise from many different causes. So we can have really rapid
changes in the driver so the driver changes rapidly in the system responds. We can have changes in disturbance regimes like fires, hurricanes, floods.
The types of things have been very much in the non-fake news this year. We
can have stochastic variability or increased patterns or changes in
variance in drivers even in the absence of a changing mean. In real world systems,
we have multiple drivers interacting with each other
changing simultaneously so it’s really hard to disentangle what happens in the
end. Thresholds are particularly thorny because it’s really hard to anticipate
them before you pass them and see that you go down that slope and finally, the
theory. We have a lot of theory to address these things but the theory has
really outpaced the real-world applications. When you try to figure out
how do I take some of these ideas that are out there and apply them in the
system in which I work.So, I’m going to be telling you a story about fire,
climate, and forest resilience and a story about abrupt change so I just want
to put a little context for when I talk about disturbances and drivers what I
mean. So, if we start here and we have a system or state or any any of
whatever your favorite response might be we may have disturbance events that
happen in time, they happen to be equally spaced here this is just a cartoon. So,
they affect the system, the system recovers. They affect the system, the system recovers. This is
kind of like the normal state of being that the system is accustomed to. However,
there are ways in which the changes in these these dynamics or their
interactions may lead to situations where the system doesn’t recover. So, maybe the disturbances become so frequent in time that the system can’t
recover and maybe it crashes. Maybe the intensity of the disturbance
or the size of the disturbance is increasing over time and again causes a
disruption of the disturbance recovery cycle and then we have may have changing
drivers interacting with disturbances that again can operate to tip the system. So, conceptually that’s where I’m thinking but I’m going to take you on a
trip to Greater Yellowstone today so out of Georgia and off to the western US. This is a place that I’ve worked in for almost 30 years now. It is one of the
largest intact wildland-ecosystems in the temperate world. It’s located in the
northwest-corner of Wyoming, centered on Yellowstone National Park, surrounded by
other parks and wilderness areas. I also want to just acknowledge here briefly
that I’ll be talking about work that is done in collaboration with my students
and colleagues and funded by a variety of different sources. So, Yellowstone is
really well-known to most of you. How many of you have been there? Okay so all of you. How many of you that have been there and remember seeing blackened trees?
oh good okay so handful of you or I’d say maybe half of you have been. So, most
people go out there and they’re familiar with the wildlife the thermal
features and the scenery and all of that. Yellowstone is dominated primarily by
forests. It’s about 80 percent forested. Most of those forests are middle
elevation dominated by lodgepole pine. There are spruce fir forests at the
higher elevations that are cooler in and at the lower elevations that are
warmer and drier we have Douglas fir and Aspen. Now, still
a little hard to see in here with the lights, but in the summer of 1988,
Yellowstone had very very big wildfires and these fires were the ones that were
on the news every night, much as California has been throughout the fall.
At the end, the forests looked like this. And as I like to say, this picture I took
this in October of ’88 it is a color picture just looking like it’s black and
white. So, it looked like the area was quite
devastated, however, we have learned and I’m going to walk you through a little
bit of the foundational work because I need to establish that before talking
about some of the change. So, I’ve been studying these fires since 1988 and I’m
just going to be do a little bit of a whirlwind about what we have learned
since then and then how we’re thinking as we look ahead. So, is actually not
new in the system and I’ve had fun reading some of the early reports of the
journals of the explorers that were surveying that country for the first
time and in one quotation here from Langford who ended up being the first
superintendent of Yellowstone. They talked about going through breaking camp,
traveling along the edge of the Firehole River, passing through a long stretch of
fallen timber blackened by fire for about four miles so my colleague Bill
Romme has done the dendro work and the fire history reconstructions in
Yellowstone and this was likely the 1862 fire that happened right in that area
that is reported. So, fire has been there even since euro-american settlement,
however, since 1988 we’ve also learned that infrequent, stand-replacing fires
meaning the fire comes through burns through the canopy of the trees kills
the trees this is very different than the coastal longleaf pine or loblolly
pine types of systems. These kinds of fires are
business and as usual so throughout the past 10,000 years so throughout the
Holocene fires have recurred at one to three hundred year intervals some
variation over time in that system. Importantly they’re driven by climate,
not by fuels. So, fuels are generally always available but most summers have
climates that are too cool and too wet to burn. So, when you get that infrequent
summer like 1988, it makes all the fuel available and the fire just continues
goes through the landscape and resets the system. So, the 1988 fires are notable
in the West actually in in fire ecology for sort of ushering in the new normal
or the new era of wildfire in the west in which we are living now each year. So
nonetheless the size and the severity of the 1988 fires were really surprising to
scientists and managers alike because we hadn’t seen others that big and and that
severe during the 20th century. Largely side comment, because the climate was too
cool in too wet we didn’t have the burning conditions happen. These fires
burned under very severe droughts, very high wind and I always have to mention
they were not caused by past fire suppression. This is a different system
than the southwestern ponderosa pine or even some of the savanna type settings
that we have in the east. And for a budding landscape ecologist, they gave a
wonderful opportunity to study a landscape scale creator of pattern that
we can’t do experimentally. Shown here is the outline of Yellowstone. The red areas
are the perimeters of the fires so they affected about a little bit more than a
third- almost 40% of Yellowstone. So, I got to go up in a helicopter in October of
1988 which is really fun it was when they were still firefighting and I was
out there trying to figure out what the research would be and what
things look like and get the hypotheses together for an NSF proposal so I’m
gonna again do a quick run through some of the main messages from 25 years of
work. One is that the fires created a really complex landscape mosaic. We’re
accustomed to this now any time you have a big disturbance you know it’s gonna be
heterogeneous. We didn’t know that at the time, which sounds kind of silly but
flying over in the helicopter and seeing this kind of pattern. Where we have
different patches of highly, very severely burned areas where all the
trees are killed, islands of green trees that are missed by the fires, and then
these brown perimeters here where you have the trees killed but they didn’t
burn off for the needles. So, very complex spatial mosaic which in turn influenced
successional patterns. We also were surprised that the vegetation recovered
really rapidly. Like much more rapidly than we expected, and if you — the sequence
this is October in ’88, this is two years later you can see the flowering, the
robust flowering and the understory vegetation recovering. By 15 years, these
are all the little lodgepole pine trees and they start out by recruiting the
very first year after fire. Little baby seedlings but they come in really early. The understory vegetation re-sprouted primarily and that was also a surprise
so the fires didn’t burn down into the soil very deeply which we were surprised
at initially, and so in many cases you can see here these are lupins you can
see this is a very well-developed root that the plant has restarted from. So, we
did a lot of excavating at the time but basically in 89, the plants re-sprouted. In 1990, two years after fire, they flowered and in 1991, we had a huge
recruitment of seedlings of native flora. So, non native plants did not increase
which was not what we expected based on what was known previously. In general, species richness at a plot level increased for
about five years or so this is from widely dispersed plots around the park. There was a still strong influence of the abiotic template so the composition
was kind of similar initially following the fires but then it diverged based on
the elevation, the topography, the soils and the like and there was a strong
effect of ecological memory. So, the species that came back were largely the
species that were present before the fire because of all these autogenic
mechanisms. The abundance of the lodgepole pines was, I would just say,
astonishing. So, they come back really fast but the variation was so much more
than we ever expected. Everything from a sparse forest coming back so that’s
about 500 or so trees per hectare to over 400,000 actually over 500,000 stems
per hectare. I don’t know how — for those of you who aren’t familiar with
that means right now we can’t walk through them because they’re so dense
then just extremely densely packed. So, this is all the same stand age coming
back after the same event but very very big differences in the structure in
the ecosystem. Largely due to variation in whether the
trees bore serratinous cones those are the kinds of cones that remain closed until
they’re heated and then they release their seeds but that trait varies across
the landscape in ways that we didn’t know. We kind of serendipitously spanned
the gradient and then also this variation in fire severity had an
influence on these patterns. So, when we most recently re-sampled these at 24 and 25 years after the fire, we still have this variation on the landscape. It’s
kind of what it looks like it’s some really challenging sampling conditions
right now because the trees are really dense. You can’t you can’t run a bearing. You can’t sight very far and all of the pre-fire coarse wood, the standing snags, standing dead they’ve all fallen. So, you have standing down up to your
nose trees that you can’t see through and you’re trying to swim your way and
climb your way through to do your sampling. It’s a really good place to
send undergraduates of graduate students So, these trees are also really
productive. Their actually their net primary productivity is higher
than the mature stands at this point. The numbers here that are showing for
averages for those of you who know- think in those terms are there. But suffices
to say they are really really productive and even in those high
density areas where you’d think they’d be out competing each other already,
they’re not. So, the quantity of trees at the ecosystem-level trumps the quality
of the trees at the tree level so if you have three or four hundred thousand
trees per hectare as is shown here even though each individual tree is smaller
than where they’re they’re open grown at the ecosystem level that is huge amounts
of productivity. These patterns of differences in stand structure set up
a pattern across the landscape, a mosaic of process rates. And so this is
ten years post-fire. The patterns are still similar but there is variation in
the total above-ground net primary productivity and we have a mosaic. This
is the southern portion of Yellow Stone of process rates that’s due to the
patterns that were set up following the disturbance. We also know that that mosaic persists for over 150 years. This is work by one
of my former PhD students Dan Kashian just showing you that the density of
trees per hectare goes down over time so by the time you’re at 200 years
following a fire, it’s about 1,200 trees per hectare but this is the coefficient
of variation among stands of the same age so the variation remains high and
then it settles in by about 200 So, those initial patterns really set the stage
for how the system behaves for a very long time. I also want to talk briefly about nitrogen dynamics. I avoided that with great success while I was a
graduate student I did not do any biochemistry and I got really interested
in it in the 1990s because wondering what does this mean for the function of
these systems. So, started doing it then but surprisingly there wasn’t very much
known about nitrogen cycling following this kind of fire. Almost everything had
been done on prescribed, low-severity fires. So, we did a bunch of stuff on it
and two main points. One is that despite what you have learned from the
Hubbard Brook examples and what we all teach in undergraduate ecology, this
system did not lose much nitrogen following the big fires of 1988. We kept
thinking our data were wrong. Okay going back, going back and back back again. But the microbial community in the soils is really tight it’s holding on to the
nitrogen that’s remaining in that system and we know that from laboratory
incubations using pool dilution. Where the consumption the
grabbing by the microbes exceeds what’s the net production and also from field
incubations that are year-long resin core incubations. So, the microbes in the
system really hold on to the nitrogen. By about fifteen years, the plants those
rapidly growing lodgepole pines that I showed you they’re really effective at
competing with the microbial community in the soil they’re also mycorrhizal and
they become a very strong sink for nitrogen so they start to sequester it
and then similarly to what I showed you for carbon or for productivity this
variation and tree density that I was showing you, also sets up a landscape
mosaic a foliar and and and nitrogen cycling rates. As of twenty five years, we
still have no evidence that nitrogen is limiting productivity in this system
which is very surprising and counter to what conventional wisdom would be. The
foliar nitrogen concentrations are still surprisingly high.There is a negative
relationship between productivity and nitrogen availability. That’s opposite
what you would expect if and availability was associated with
increased productivity and over time all pools of nitrogen in the system have
increased and I will say we have very few, we don’t have alder, we don’t have
like here you have black locust Seca we know we don’t have a nitrogen fixer
that’s dominant so where it’s coming from is something we’re still working on. So, basically the consequences of those big fires have been very well studied. We’ve we’ve studied them to death. I shouldn’t say that because it’s a lot of
fun to still try to track it. We know the narrative. So, this is from National
Geographic almost two years ago now We know the narrative so
we have the fires and then we have the recovery and then we have the forest
coming back just like that cycle that I showed you at the beginning. So, the
bottom line native vegetation ecosystem ecosystem processes recovered rapidly without
intervention. Yellowstone is well adapted to these kinds of fires. Thank you very
much lots of resilience. One of the ways that
we depict this in another way and I’ll use this again later is by these ball
and Cup Diagrams so if this is my Yellowstone forest, the fires can come
around and they they move that ball around in this basin but they don’t push
it out, so the system you know moves out but it comes back moves out and comes
back whether or not it can flip to another state is one of the things that
we’re starting to think about. So, we know going back to the issue of change that
both climate and our fire regimes are changing. The paper that Leroy Westerling
published in 2006 in Science was the first one to show the statistically
significant relationship between climate change in the West and the occurrence of
fires. So, we were having large – an increased number of large fires high
severity fires in the West and it’s associated with the warming temperatures,
with the earlier snowmelt in the spring, and then the lengthening of the fire
seasons so this is all stuff you’ve been hearing on the news as well associated
with California. Leroy updated this these bars here are the
numbers of fires that’re large in the west. Idec the bars of the decadal means
and you can see them marching steadily up and that’s continuing. So, we started
thinking about the effects of climate change in Yellowstone in like
really in the 1990s so The first paper Bill Romme and I wrote
came out in 1991. It’s before we had the the sophisticated climate models or the
predictions nor even really enough data yet to say that the trends were clear. So,
it was really more of a thought exercise about what would happen and so you know
we laid out that yes well a warmer and drier climate would increase fire
activity. If we had more fires it probably would reduce the net age of the stands
across the landscapes and again it might shift the vegetation upslope just
because of the cooler conditions at higher elevations. So, we did probably
over ten years or so more than that actually because too through 2009 a lot
of different modeling approaches where we really kind of I you know I put the
hammer here because we really tried to hammer the system based on all the
variability that we had seen throughout the records from the Holocene so what
happens if we make the fires as frequent as they were observed and the like. So, in
all of those cases we know that climate and fire had changed in the past that it
would change in the future. In all cases when we were doing our
modeling explorations the forests were recovering just finds it was very
consistent with what we had seen following the 1988 fires and we knew
that those were not catastrophic. So again, it did not change the conclusion
that Yellowstone is well adapted to these large severe fires and it was
likely to be in the future. Or is it? So, the thing when I talk about abrupt
change sometimes we have changes in our conceptual understanding as well and
this was a watershed for me. So, Leroy and I were at the same
conference and he put this map up of the moisture deficit in 1988, the year of the big fires and the redder, the drier, so that’s where it’s more arid with the
drought conditions are the worst and you can see in 1988 it centered right on
Yellowstone we know what happened then. And then he showed this for the
projections for 2090. Where it’s both more intense and that
red throughout the West. That is out of the box. That is beyond what we had
considered because it’s beyond what we had seen in the Holocene. It goes outside
of that range and although we had been really thinking we hadn’t thought than
anything that severe could be in our futures so it really started changing
our understanding our thinking about what might happen. So, Leroy and I were able to get with a couple of other colleagues some joint
fire project funding to start looking at what those implications might be and for
the Yellowstone area this suggested that we see spring summer temperatures up by
four to six degrees C and that’s the time period where it matters for fires
by the mid 21st or the end of this century. The water year deficit is driven
not so much by a change in precipitation but by that warming which is tending to
dry out the system. This is a bit of a complicated figure and I’ll just make a
couple of points from it but we looked at what that would mean with many like
with ten thousand replications and such from different climate models
looking at the log area of area burned on the Y and then time on the X. The area
burned in the 1988 fires is here and all of the vegetated area of Greater
Yellowstone is here. These colors are showing the observations which match
well the median and then the whole full range of observations over the fires
over the projections. Basically, what it says is by mid-century, we’re getting
very few years with no fire and the weather conditions associated with big
fires are happening essentially all the time. It doesn’t say there will be fires
it’s based on the weather conditions because there are a few feedbacks here. So, the the nugget here is that the novel the fire regime in the future could be
novel relative to the to the Holocene to the past 10,000 years and these changes
are much greater than what had previously been considered and we would
have few years without fire. Fires would no longer be climate limited so
we would have the climate available for those big fires all the time which
remember I said it was quite different from the beginning eventually fuels
would have to be limiting to the fires and what would happen to the fire
severity remains to be seen. So, this has really sent us on a new
investigation of what might be happening throughout the West so we know again that fire activity will increase but the details about how that will play out
remain the subject of very intense research by not only our group but
others. How many fires? How much area will burn? When will these changes happen?
Where will we see them? Are there going to be tipping points that lead to
fundamental changes in those systems? So, this is hard to do because by their
nature these high severity fires are infrequent at any given place. It’s not
like you can easily go out and get lots of replication. Trees live a long time
and so there can be very long time lags involved before you see
changes but the fires themselves can potentially trigger an abrupt change in
that whole fire and recovery cycle. So, what do we do in the face of these
challenges? So, I would say for this and many of the other big, wicked problems we
face we can’t put our heads in the sand and ignore them we actually need to take
advantage of all of the sort of tools in our toolkit so observations across a
disturbance characteristics are across space, long-term study where we can look
at the dynamics as they change, experiments, and then also process based
models. It’s going to be really important I think that we try really hard to
identify the nonlinearities or the thresholds that might be associated with
abrupt change and that we understand the mechanisms or we test hypotheses about
the mechanisms that could be underpinning such changes so we’ve been
doing that in the Yellowstone system asking how does the warming temperature
plus the changing fire, going to effect the forests in the future and I’m going
to walk you through three different mechanistic hypotheses about what might
play out. The increase in fire frequency affecting the supply of seeds, the fire
size affecting the delivery of seeds into burned area, and the drought affecting establishment. So, first of all frequency
of fire and seed supplies, so we have conifers they’re obligate seeders, they
have to produce cones that’s the source of the seed that comes in whether
they’re serotinous or non-serotinous. If the re-burns occur before the trees have
matured, you lose your seed source, and so that could lead to a failure of the
ability of the system to recover, so following the 1988 fires, it’s mostly
large trees big mature trees long interval fire that set up this variation
nonetheless but still a lot of recovery following the fires. However, we are now
starting to see re-burns of those areas that burned in 1988. So, in 2016, it was
not in the news because there were fires burning in the West everywhere else that
were more– of a greater threat But we had the most area burned in Yellowstone
since the 1988 fires. This picture shows the fires of 2016 burning in some of my study areas the 28 year old lodgepole pine stands, so we were just out there a
year ago sampling with with rapid funding from NSF and this is what some
of these look like and again it’s a little bright, so you can still see there
is this mosaic just like I showed you from my helicopter picture but we also
have areas like this we were calling them “stump towns” because all of the
coarse wood and all of the young trees were consumed in some of these places. It’s greater severity than we had seen following ’88, so in areas, this is I
haven’t even analyzed the data, this is really the back of the envelope stuff
but in places where we had kind of like the normal stand replacing fire, the
trees that came back after the ’88 fires this density is matched by what’s coming
back after 2016, however, in the areas that look like this I mean they’re just
remarkable you can these lines here these are ghost logs so
that’s where the coarse wood had been on the ground but it was completely
consumed and you can see there’s no above-ground trees. We had to
use a stick to poke to try to recreate what the densities were because all that
was left with stumps. We have there 99% reduction in the regeneration so these-
we really did lose the seed sources so in addition to the frequency this
variation in severity is also playing a role I think in what will come back. In
terms of fire size if we change the patch sizes and for species like Douglas
fir that do not have a canopy seed bank then fire size may influence whether or
not seeds can disperse into these areas. This is post ’88 fire data but here’s
surviving Douglas fir trees and then we have coming down the hillside here
little post ’88 ones coming back in but essentially if you’re more than 100
meters from a live tree, a seed source there’s very little regeneration and
especially if you’re on a dry position. So, patch size will matter and then
drought is associated with the fire but drought can also affect the ability of
the trees to establish and grow following a fire. So, if we think about
the mature forest big trees can handle a lot; little trees not so much. Just like
in your garden when you’re planting flowers or vegetables you’ve got to baby
them in the beginning. So, the tolerance for the mature trees of environmental
conditions is much broader than it is for seedlings. So, we’ve been looking, we
took advantage, well again one of my former PhD students Brian Harvey, we
looked at fires throughout the Northern Rockies that were followed by three
years of above average temperature so lower moisture and then normal or wetter
conditions and we found that indeed we see fewer trees in the dry post fire
years and also on south-facing exposures where it would be would be drier. So, there’s evidence for all of these mechanisms coming into play. We’re also
doing experiments. One of my current students, Winslow Hansen, we are in the
process of writing up this paper for ecological monographs but we’ve done
both greenhouse experiments and field transplant experiments where we’re
growing seeds and post-fire soils of current climate and then in places in
the landscape today where the future projected climate is is apparent today
some of the lower elevations and those data are also showing that at the low
elevations over four years field study we have much reduced success in terms of
tree establishment. So, we’re getting this from from multiple angles. So, my question
is then are we potentially seeing forests in transition where instead of
this historical condition that I showed you we may be making this Basin more
shallow by warming the climate and then with the changes in the fire have the
ability to perhaps change the system to a non forest or deciduous forest State? So what that would look like is do we have Yellowstone transitioning from this
landscape which is what you see now if you’re out there to something that has a
whole lot more open type vegetation maybe an expansion of some of the lower
elevation grasslands expansion of Aspen and of douglas-fir. So, one of the
challenges even with these kinds of field studies is we kind of get what
we’re given by the weather and the conditions that we have but when we’re
trying to understand the suite of factors that could affect novel
conditions in the future in novel systems it’s it remains challenging and
when we go to use models that exist that are empirically based they’re based on
the past not on the future so we may be seeing climate conditions and fire
regimes that are quite different from what we have seen in the past and we
want to know how multiple interacting driver
can shift us to different different positions so therein comes the role for
modeling. So, we are now using a model called “Iland” developed by my
collaborator Rupert Seidl who’s in Vienna which is a process based model,
individual based model but scalable that represents trees and landscapes and
disturbances and spatial dynamics and such so I just want to mention what it
is I’m not going to go through the details. One of my students has already
parameterize this model for our Yellowstone tree species. This is just showing
in the gray lines actual simulations from our stands based on the ’88 fires. In
the red dots showing field data that we have from plots so we have the model
behaving well and then I mentioned these factors. We would like to look at how
they interact with one another and so we have conducted a factorial experiment
with the model looking at several two species but two forms of one fire
return interval from the lowest observed in the Holocene to shortest that we’ve
observed in the field, varying distances from seed source, and then climate
periods that are the historical mid 21st century and late 21st century and looked
at all the combinations of those and I’ll just show you one output here
that’s all this paper is being revised now we have minor revision for ecology
so hopefully it’ll come out year. But for each of our species what
I’m showing here is a state space of where in red post-fire regeneration has
failed and we were very conservative about that that’s less than 50 trees per
hectare so that means you’re really you’re not even in a savanna
by that point and then where it’s successful as a function of distance to
seed source the return interval of fires and the climate periods. So, you can see
the various combinations that give you the possibility for having a non forest
coming in at the end. So, in all cases distance to seed source matters a lot
fire return interval is especially important if you have an aerial seed
bank and the like so we’re trying to explore what sets of conditions might be
there and then we’ll look for where those might happen on the landscape. So, the bottom line for this I think that affects all of us no matter what systems
that we’re working in is that what we’ve observed in the past even the deep past
even if we go back ten thousand years can inform us about mechanisms but it
may not be enough to help predict the future In my case, I think forest state
may be less resilient in the future as we have climate warming and we have
change in fire regimes I always like to make sure that I’m not really sounding
like the sayer of doom. Yellowstone is not going to be destroyed it will still have
native species it is still one of the best places on earth to observe how
nature is responding to the changes with minimal human impact, additional. We will
see changes in the age the type the extent location of forests and I think
by focusing on mechanisms particularly when we have long-lived species and time
lags involves that may help us with early detection and again following my
theme here I think anticipating when where and why we’re likely to have these
big abrupt changes is a challenge that I think we should all be considering in
our systems. So, I’m going to end this in my last slide here but a couple of extraneous comments that are geared particularly towards the early career
people who are in the room and I think we have- these are some of the lasting
lessons I got as a student here that have helped me I think throughout my
career but some of them are are looking forward a little bit. One is always have
good questions. You can push to ask a good question regardless of the system,
regardless if it’s applied or basic research or any of that and follow the
things that are really exciting to you. I mean you should want to get up in the
morning and go to work and figure out what new whatever the problem is you’re
working on so really really push on the questions. I think also don’t get
caught up and having just one one tool like you see every hammer — every problem
is a nail because all you have is a hammer you have the opportunity to hear,
to learn diverse approaches use them and appreciate them in your work so whether
it’s experiments comparisons long-term data, modeling, remote sensing, the list
goes on however in the world of big data and fancy statistics that some of us
have to catch up on and learn as we go retain your close association with the
real world system if you’re modeling or doing statistics your work will be
better informed by knowing your system. If you’re down in the weeds with your
system you may be also better informed by taking the broader modeling approach but
use them together but don’t lose the connection and then finally in the world
of fake news which we live in right now which is very very disturbing two things
that I think are really important with the advance of the advent of predatory
journals and the like strongly support the value of peer-reviewed science the
system is not perfect but I don’t think we have a better one and then also get
involved in your scientific society support the professional organizations
that are advocating on behalf of all of us and all of our science. You can see I’m passionate about it so with that note I will say thank you very much.
I really am honored to have had the chance to attend and talk to you. Thank
you! [applause]